ISS Accelerator Working Group

1
ISS Accelerator Working Group Summary and Future
Plans R.C. Fernow BNL ISS Meeting KEK 25
January 2006
2
Outline

• proton driver
• targetry and p production
• front end
• muon acceleration
• storage ring

• I will concentrate here on recent activities
• a comprehensive overview was given by Chris
  Prior at CERN
• his talk includes many topics that are not
  covered here

3
Proton driver overview

• which PD type is best suited to meet neutrino
  factory requirements?
• some important issues
• beam current limitations
• creation of short bunch
• repetition rate limitations
• space charge
• tolerances
• PD design tied to downstream constraints (pulse
  structure frequency)
• most PD designs based on lab-specific,
  multipurpose machines
• some of the new designs use FFAGs

4
Proton driver designs

• SPL 3.5 GeV 50 Hz 4 MW
• JPARC 50 GeV 0.33 Hz 0.6 -gt 4 MW
• AGS 24 GeV 0.5 Hz 0.2 -gt 4
  MW
• FNAL SCL 8 GeV 10 Hz 0.5 -gt 2 MW
• RAL RCS 5 GeV 50 Hz 4 MW
• RAL RCS 15 GeV 25 Hz 4 MW
• RAL FFAG 10 GeV 50 Hz 4 MW
• KEK/Kyoto 3 GeV 1 kHz 1 MW
• A. Ruggiero FFAG 12 GeV 100 Hz 18 MW
• A. Ruggiero FFAG 1 GeV 1 kHz 10 MW
• .........

5
JPARC

• initial linac 0.18 GeV
• uses 3 GeV booster
• upgrade power by
• increasing rep rate
• main ring commissioning
• in May 2008

(M. Tomizawa, Y. Mori)
6
CERN SPL

• major upgrade of p injector complex SB, ßB, NF
• 2x beam brightness, improved reliability
• 3 MeV test stand in 2008, 160 MeV Linac4 in 2010
• uses 704 MHz bulk-Nb cavities

• more work needs to be done to determine
• if it can meet NF requirements

7
AGS Upgrade to 1-4 MW
Alternative scheme with 1.5 GeV FFAG injector to
AGS

• 1.2 GeV superconducting linac extension for
  direct injection of 1 ? 1014 protons low beam
  loss at injection high repetition rate
  possible further upgrade to 1.5 GeV and 2 ? 1014
  protons per pulse possible (x 2 power)
• 2.5 Hz AGS repetition rate triple existing
  main magnet power supply and magnet current
  feeds double rf power and accelerating
  gradient further upgrade to 5 Hz possible (x 2
  power)

8
FNAL 8 GeV linac

• stripping of H- is an issue
• blackbody, magnetic, residual gas
• needs RF amplitude phase correction
• before injection into MI, 1.5 km?
• needs accumulator ring for NF

9
RAL 4 MW FFAG Proton Driver

• RAL design
• 5 bunches per pulse
• 50 Hz repetition rate
• 10 GeV
• Isochronous FFAG with insertions

• RF system naturally gives 2 ns rms pulse
• need to add 6th harmonic to get 1 ns rms

(Chris Prior)
10
Proton driver plans

• identify issues for producing short (1-3 ns
  rms) bunches
• define parameters for bunch compression ring or
  transport line
• evaluate space charge issues
• look at implications of pulse structure

11
Targetry and p production overview

• what is the optimum target material for 4 MW?
• what constraints limit target operation at 4 MW?
• e.g. proton bunch intensity, repetition
  rate
• driver/target/capture systems need to be jointly
  optimized
• to get the maximum number of neutrinos at
  the detectors
• new pion production/collection simulations
  (Brooks, Kirk)

12
Liquid target issues

• liquids may be only practical choice above 2 MW
• mercury, molten metals
• MERIT experiment at CERN will demonstrate
  feasibility of mercury jet target
• uses parameters relevant to neutrino factory
  targets
• instantaneous energy deposition corresponding to
  4 MW beam
• 20 m/s jet velocity
• 15 T solenoid field (US, KEK) or 0 T (CERN)
• active simulation effort to understand jet
  dynamics

13

MERIT experiment at CERN

• MERIT (MERcury Intense Target )
• PS 24 GeV beam, 2.8 1013 protons on 1.2 mm ? 1.2
  mm beam spot
• Peak energy deposition 180 J/g
• Beam on target April 2007

14

MERIT recent developments

• project reviewed on 12 December 2005
• pulsed solenoid completed and delivered to MIT
• demonstrated rad-hard optical diagnostic fibers
  at CERN
• angle of jet to solenoid changed to 33 mrad to
  give same jet aspect ratio as NF
• set operating temperature at 80 K (reduced cost)
• good agreement between experiment ANSYS on
  Ti-Al-V beam windows

15 T pulsed solenoid in cryostat
Mercury jet delivery system
15

Solid target issues

• shock is a major issue
• radial longitudinal stress waves
• need to understand material properties under
  irradiation
• e.g. strength, thermal expansion changes
• need to control heat flow from target
• forced He cooling
• radiation cooling (levitated ring)
• providing new material for each proton pulse
• e.g. rotating band, bullets

16
p production on Ta
S. Brooks using MARS15 after RAL phase rotation
Doubled lines give some idea of stat. errors
Optimum moves down because higher energies
produce pions with momenta too high for capture
17
Pion production on mercury
(H. Kirk)

after US Study 2a cooling channel
18
Targetry and p production plans

• assess minimum acceptable proton beam rep rate
  at 4 MW
• evaluate possibility of realistic solid targets
• e.g. rod, band, pellet, granular
• continue study of p production for
  intermediate-Z targets
• incorporate HARP/MIPP results for p production

19
Front end overview

• front end p collection, bunching, phase
  rotation, cooling
• comparison of existing designs
• US study 2b, CERN, KEK
• examine new FE ideas
• improve theoretical understanding
• experimental program examining important RD
  issues

20
MUCOOL RD

• MUCOOL design, prototype and test
• cooling channel components
• Muon Test Area completed at FNAL
• 805 MHz cavity installed
• high power testing underway
• 201 MHz cavity delivered
• final hook up underway
• 400 MeV p beamline has been designed
• RD program underway to understand
• RF breakdown
• effect of magnetic field on maximum
• gradient is important design issue

(A. Bross)
21
PRISM update
(A. Sato)

• Phase-rotated intense slow µ source
• FFAG ring under construction
• designing injection line detector
• FFAG ring commissioning in 2007
• RF cavity achieved 170 kV/m at 5 MHz
• studying vertical injection/extraction
• doing detailed tracking now
• dynamic aperture
• backgrounds for physics

22
MICE

• ionization cooling channel demonstration
  experiment at RAL
• first beam to step I of experiment in April
  2007
• tested cryostat for hydrogen absorber
• SciFi tracker test at KEK was successful
• plan to test MICE production target in ISIS June
  2006
• design completed for spectrometer solenoids

(M. Yoshida)
23
Beam-target survey

• positive muons
• using standard beam files from Stephen Brooks
• cf. US Study 2a had 0.007 µA / p / GeV

24
Effect of proton bunch length

• it is important for proton driver to deliver a
  short pulse 3 ns rms
• performance degrades rapidly for longer pulses
• easier to phase rotate higher energy particles
  if pulse 1 ns rms

(J. Gallardo, H. Kirk)
(US Study 2a, 24 GeV p on Hg)
25
1D model of phase rotation

• try to better understand optimum parameter
  choice for variable
  frequency phase rotation
• shorter rotator doesnt have enough integrated
  gradient
• lose higher energy tracks
• phase rotation works better if reference
  particle momenta
• are changed adiabatically

(R. Palmer)
26
New configurations
(D. Neuffer)

• cool while doing phase rotation
• cost savings
• 150 atm hydrogen (room temp)
• 24 MV/m RF
• performance looks promising

(A. Klier)

• Guggenheim cooling channel
• provides longitudinal cooling
• solves problems with
• injection, absorber heating
• can taper parameters

27
Cooling versus accelerator acceptance

• can we save money by reducing amount of cooling
  and increasing accelerator acceptance?
• present work depends on cost model for
  non-scaling FFAGs by J.S. Berg
• still needs more work on unresolved issues
• cost modeling for other accelerator systems
• effects of FFAG acceptance loss with AT
• trade-off between amount of cooling and detector
  size

28
Front end plans

• continue studying
• acceptable design for the KEK 0.3-1 GeV/c ring
• cooling vs. accelerator acceptance
• optimized phase rotation
• combined cooling phase rotation
• effects of short proton pulse lengths (lt 5 ns
  rms)
• check suitability of existing absorber and
  window designs
• 4 MW, 2 signs
• monitor developments at MuCool, MICE and PRISM

29
Muon acceleration overview

• compare different acceleration schemes
• reference designs NuFactJ, CERN, US Study 2b
• new design RAL FFAG
• some issues
• transverse longitudinal acceptance
• beam dynamics during acceleration
• implications of keeping both sign muons
• matching between acceleration subsystems

30
Reference µ accelerator designs
NuFactJ
CERN
US Study 2b
31
RAL FFAG µ accelerator
linac
RLA
IFFAG
RAL design

• linac from 0.2 1 GeV
• RLA from 1 to 3.2 GeV
• two isochronous FFAGs from 3.2 to 8 GeV
• and 8 to 20 GeV in the same tunnel.
• IFFAG can use any RF frequency,
• constant QV gt easier collimation

32
Scaling FFAG issues

• tunes are constant, avoids resonance crossings
  during acceleration
• constraint that rings fit inside JPARC leads to
  non-optimal design (J.S. Berg)
• costs of present design are likely to be high
• need tracking with soft-edge magnets for a green
  site
• need to specify parameters for RF system
• examine longitudinal dynamics
• need to look at spiral sector FFAG designs also

33
Machine parameters
150 MeV FFAG experimental results
(M. Aiba, Y. Mori)

• have overcome close orbit distortion
• have overcome resonance crossing (COD)
• extracted beam from FFAG
• for first time
• 100 Hz operation successful after RF
  improvements

34
Non-scaling FFAG issues

• non-scaling FFAG can minimize orbit excursion
• beam dynamics issues
• longitudinal acceleration outside RF bucket
• transverse tunes are not constant,
• crosses structure resonances
• longitudinal behaviour coupled to transverse
  amplitude AT
• some particles with large AT are not accelerated
• for more moderate AT, get longitudinal emittance
  blowup
• time of flight vs. energy depends on transverse
  amplitude

(S. Machida)

• tune per ring
• with alignment
• errors
• leads to emittance
• growth

35
Effect of transverse amplitude
(S. Machida)
(F. Lemeut)

• S. Machida tracking code
• injects distorted beam from 10 GeV
• ring into 20 GeV ring
• sees large ?p after 2nd ring
• also sees distortions in transverse
• phase space

• 6D multiturn tracking with Zgoubi
• 5-10 GeV KEK FFAG ring
• looked for ellipse orientation
• aspect ratio that minimizes
• longitudinal distortion

36
US Study 2b RLA optics

• symmetric dogbone-RLA accelerates both µ and µ-
• 3.5-pass (1.5 5 GeV) scheme
• linear optics design has been completed
• tolerable phase slippage in the higher pass
  linacs
• magnet misalignment error analysis shows
  manageable level
• of orbit distortion for 1 mm magnet
  misalignment error
• examined focusing error tolerance for quad
  fields (0.2)

arc 3 optics
(A. Bogacz)
37
Muon acceleration plans

• need to quantify effect of AT on NS-FFAG
  transmission
• study sensitivity of FFAGs to errors
• implications of various pulse structures
• kicker and injection/extraction designs for
  rings
• explore dogbone RLA in further detail
• clarify choice of FFAG lattice
  (doublets/triplets..?)
• program of full 6D simulations.

38
Storage ring overview

• two new designs C. Johnstone, G. Rees
• matrix of machine/detector angles C. Prior
• important design issues
• racetrack or triangle geometry
• 20 GeV or 50 GeV or 20 GeV-upgradable
• how to handle both muon charges (1 ring or 2)
• length of µ bunch train (constrains
  circumference)
• RF to maintain bunch structure
• beam loading (MW µ beams)
• shielding from µ decays

39
Johnstone design

• racetrack geometry, C 1371 m
• both 20 GeV and 50 GeV use same lattice
• (magnets run at reduced strength for
  20 GeV)
• production straight
• 496 m long (36 production efficiency)
• quad focusing
• maximum beta 155 m, 167 m
• rms divergence 0.12 / ? at 20GeV, 0.19 / ? at
  50GeV
• arcs
• maximum beta 16 m
• 6.4 T dipole fields
• uses sextupole correctors
• tracking being done by Francois Meot

40
Rees design changes

• look at isosceles triangle geometry
• design for MW intensities
• introduce µ beam loss collection
• use combined function magnets in arcs to increase
  intermagnet spacing
• use solenoid focusing in two production straights
  to minimize beam size
• use matching section bends for dispersion
  suppression
• must make small lattice changes when upgrading 20
  -gt 50 GeV

41
New Rees design

• circumference 1170 m
• length of each production straight 301 m (2
  x 0.26 efficiency)
• 5 bunch trains per cycle
• angles of triangle depend on ring and target
  sites
• designed for apex angle 22.4o
• 27, 34.5, 45, 52 also possible (1o)
• uses separate rings for µ and µ-
• rms divergence 0.10 / ? at 20GeV, 0.12 / ? at
  50GeV
• difficulty needs big injection system

42
Storage ring plans

• develop isosceles triangle ring with 40o apex
  angle
• develop strawman detector sites for triangle
  ring
• begin tracking racetrack ring (with errors)
• begin tracking isosceles triangle ring (with
  errors)
• look at other systems
• e.g. injection, abort, chromatic corrections

43
Accelerator Group Summary

• challenge is to try to reach consensus on a
  single optimized neutrino factory
  scheme
• leaning towards
• 8-12 GeV proton driver with 1-3 ns rms pulse
  length
• FFAG µ acceleration
• triangle storage ring
• working to identify critical questions that need
  additional RD
• trying to have reasonable scheme ready for World
  Design Study